the social contract that allows the web to exist at all is literally collapsing in real time
https://people.kernel.org/monsieuricon/creepy-crawlies
it's tempting to read this as a complaint about the inconvenience of dealing with slopbots from an operator of a single site, one whose beneficiary recently decided to go full slop themselves, so, like, whatever.
and yeah there's a certain amount of irony here. but…
this is not really about a single site. it's about a universal concern, one that is plaguing any medium-sized site at all. and the arms race as detailed here is disturbing. I've been predicting that Anubis will collapse for a few months now, and I was already late to that party (c.f. https://mastodon.social/@glyph/116908213902870864). And that's what this article shows: Anubis is on the trajectory of collapse, literally 33% of the way there. That was 1 year ago. Generously, we have 1 year left.
fail2ban doesn't work, anubis is about to stop working, and the companies doing this are more likely to be fly-by-night Chinese operations than OpenAI or Anthropic. (That's not to say that OpenAI or Anthropic are not doing scummy and criminal things in search of data, but they're not doing _this_.) Which means that the financial collapse of the US AI bubble is unlikely to make a dent, even if it happens soon.
A _lot_ of websites are about to go dark.
It's kind of standard at this juncture for people to say that there is no technical solution to a social problem, but the social solutions (regulation and/or burning down the data centers) sure do not seem to be right around the corner, and we need a stopgap right now.
Maybe we all add a bunch of links to Nepenthes instances all over our sites, in places humans would never click.
@glyph you can't really know who's doing it, not when it's so significantly residential proxies these days, to assume openai or anthropic aren't doing 'this', is um, very wrong at best